VOGONS


Reply 20 of 37, by carlostex

User metadata
Rank l33t
Rank
l33t
darry wrote on 2020-07-11, 01:12:
I think you may be at least partly mistaken . I can't speak for the Cancun FX, as I have never seen a photo with the sticker rem […]
Show full quote
carlostex wrote on 2020-07-10, 23:59:
AFAIK, the GS ROM Roland sued over isn't the exact same the one that came with the Cancun FX. It follows the GS standard but do […]
Show full quote
darry wrote on 2020-07-10, 23:04:

The Yucatan FX (resurrected Cancun FX) mentioned earlier in this thread uses a Dream chip and the infamous GS ROM that Roland sued over and then licensed .

I have a Monster MIDI which uses the same Dream chip and ROM, so sounds essentially the same as the Yucatan FX should . I do not find it that close to a Sound Canvas . I prefer my SC-88VL in SC-55 mode to it. I will probably be buying an X2 when they are back in stock and putting the Monster MIDI back in storage .

AFAIK, the GS ROM Roland sued over isn't the exact same the one that came with the Cancun FX.
It follows the GS standard but doesn't sound close because it had to be changed significantly.

Any user with a card that has the SAM 9407 (EWS64, Guillemot) can try the 4MB soundfont, that sounds virtually the same as a Sound Canvas. I don't remember the file name exactly now but its no wonder Roland sued.

So they had to license and change significantly which resulted in the Cancun FX.

I think you may be at least partly mistaken . I can't speak for the Cancun FX, as I have never seen a photo with the sticker removed from the ROM. But even one of Dream's datasheets confirms the Roland source sample ROM was being offered as an option alongside the CleanWave ROM .

According to ftp://retronn.de/docs/pc_hardware/wavetables/dream/9703.pdf , the ROM part numbers are as follows .

Validation for following ROMs
•GMS960800 Roland 8Mb
•GMS970800 CleanWave8
•GMS963200 Roland 32Mb
•GMS973202 CleanWave32Mb

If you look at the ROM that will be used for the Yucatan FX : Re: New Waveblaster card : Reptile Paradise - Yucatan FX and Re: New Waveblaster card : Reptile Paradise - Yucatan FX you can see it is a GMS963200

If you look at the ROM from the Monster MIDI that I have : Re: Roland SCB-55 in 2020 - fair price and how to find one? you can see it is a GMS963200

According to http://www.os2museum.com/wp/of-g-men-and-farmers/ , the Roland sampled version continued to be offered as an option along with the CleanWave version, as mentioned before .

The SAM9407 compatible GS soundbank may or may not be the same as the one in the GMS963200 ROM . It would be nice to be sure .

Can you provide some references regarding this ?

You might be right about the ROM versions and i might be wrong about what triggered the lawsuit. Also the cards i was talking about have no ROM's in them but RAM which the software loads the soundbank into when the card is initialized.

As i find the Dream GS ROMs significantly different from a real SC-55 i get puzzles why Roland would get upset about devices on markets they did not care about.

However the Dream SAM 9407 with the particular 4MB soundfont sounds incredibly close to a SC-55, in a blind test i wouldn't be able to distinguish them, as i tested a Roland CM-300 vs a Guillemot Maxi Sound Lite with the 4MB soundbank and personally i was astonished. It seems almost like the SC-55 ROM was reverse engineered and that soundfont was built from that. I don't have proof it happened this way, but i never heard a soundbank that sounded as close to a Sound Canvas like this one does. Far, far superior to a Cancun FX. In fact i also own a Terratec Maestro 32/96 whose ROM combinations sounds pretty much the same as a Cancun.

I don't have an easy way to make recordings but i can identify the soundbank and let everyone know which one it is.

EDIT: OK so i took a look and found that GSSBK320.94B and BANK4M.94B are exactly the same file. Ran a binary comparison on them and they match. The clue was that both had the exact same file header.
Now, this soundbank as played by the SAM 9407 to me sounds different from the Cancun, and for the better. It also sounds pretty much the same as my CM300.

Reply 21 of 37, by darry

User metadata
Rank l33t++
Rank
l33t++
carlostex wrote on 2020-07-11, 15:22:
You might be right about the ROM versions and i might be wrong about what triggered the lawsuit. Also the cards i was talking ab […]
Show full quote
darry wrote on 2020-07-11, 01:12:
I think you may be at least partly mistaken . I can't speak for the Cancun FX, as I have never seen a photo with the sticker rem […]
Show full quote
carlostex wrote on 2020-07-10, 23:59:
AFAIK, the GS ROM Roland sued over isn't the exact same the one that came with the Cancun FX. It follows the GS standard but do […]
Show full quote

AFAIK, the GS ROM Roland sued over isn't the exact same the one that came with the Cancun FX.
It follows the GS standard but doesn't sound close because it had to be changed significantly.

Any user with a card that has the SAM 9407 (EWS64, Guillemot) can try the 4MB soundfont, that sounds virtually the same as a Sound Canvas. I don't remember the file name exactly now but its no wonder Roland sued.

So they had to license and change significantly which resulted in the Cancun FX.

I think you may be at least partly mistaken . I can't speak for the Cancun FX, as I have never seen a photo with the sticker removed from the ROM. But even one of Dream's datasheets confirms the Roland source sample ROM was being offered as an option alongside the CleanWave ROM .

According to ftp://retronn.de/docs/pc_hardware/wavetables/dream/9703.pdf , the ROM part numbers are as follows .

Validation for following ROMs
•GMS960800 Roland 8Mb
•GMS970800 CleanWave8
•GMS963200 Roland 32Mb
•GMS973202 CleanWave32Mb

If you look at the ROM that will be used for the Yucatan FX : Re: New Waveblaster card : Reptile Paradise - Yucatan FX and Re: New Waveblaster card : Reptile Paradise - Yucatan FX you can see it is a GMS963200

If you look at the ROM from the Monster MIDI that I have : Re: Roland SCB-55 in 2020 - fair price and how to find one? you can see it is a GMS963200

According to http://www.os2museum.com/wp/of-g-men-and-farmers/ , the Roland sampled version continued to be offered as an option along with the CleanWave version, as mentioned before .

The SAM9407 compatible GS soundbank may or may not be the same as the one in the GMS963200 ROM . It would be nice to be sure .

Can you provide some references regarding this ?

You might be right about the ROM versions and i might be wrong about what triggered the lawsuit. Also the cards i was talking about have no ROM's in them but RAM which the software loads the soundbank into when the card is initialized.

As i find the Dream GS ROMs significantly different from a real SC-55 i get puzzles why Roland would get upset about devices on markets they did not care about.

However the Dream SAM 9407 with the particular 4MB soundfont sounds incredibly close to a SC-55, in a blind test i wouldn't be able to distinguish them, as i tested a Roland CM-300 vs a Guillemot Maxi Sound Lite with the 4MB soundbank and personally i was astonished. It seems almost like the SC-55 ROM was reverse engineered and that soundfont was built from that. I don't have proof it happened this way, but i never heard a soundbank that sounded as close to a Sound Canvas like this one does. Far, far superior to a Cancun FX. In fact i also own a Terratec Maestro 32/96 whose ROM combinations sounds pretty much the same as a Cancun.

I don't have an easy way to make recordings but i can identify the soundbank and let everyone know which one it is.

I absolutely agree with you that the GS ROM is not that similar to a Sound Canvas .

I am starting to be tempted to get a SAM9407 base card with RAM just to test that soundbank you mention . Please let us know, when you get the chance, which 94B soundbank you are referring to . My guess is the GSSBK320.94B one (from reading other posts, not for having heard it) .

Reply 22 of 37, by carlostex

User metadata
Rank l33t
Rank
l33t
carlostex wrote on 2020-07-11, 15:22:
You might be right about the ROM versions and i might be wrong about what triggered the lawsuit. Also the cards i was talking ab […]
Show full quote
darry wrote on 2020-07-11, 01:12:
I think you may be at least partly mistaken . I can't speak for the Cancun FX, as I have never seen a photo with the sticker rem […]
Show full quote
carlostex wrote on 2020-07-10, 23:59:
AFAIK, the GS ROM Roland sued over isn't the exact same the one that came with the Cancun FX. It follows the GS standard but do […]
Show full quote

AFAIK, the GS ROM Roland sued over isn't the exact same the one that came with the Cancun FX.
It follows the GS standard but doesn't sound close because it had to be changed significantly.

Any user with a card that has the SAM 9407 (EWS64, Guillemot) can try the 4MB soundfont, that sounds virtually the same as a Sound Canvas. I don't remember the file name exactly now but its no wonder Roland sued.

So they had to license and change significantly which resulted in the Cancun FX.

I think you may be at least partly mistaken . I can't speak for the Cancun FX, as I have never seen a photo with the sticker removed from the ROM. But even one of Dream's datasheets confirms the Roland source sample ROM was being offered as an option alongside the CleanWave ROM .

According to ftp://retronn.de/docs/pc_hardware/wavetables/dream/9703.pdf , the ROM part numbers are as follows .

Validation for following ROMs
•GMS960800 Roland 8Mb
•GMS970800 CleanWave8
•GMS963200 Roland 32Mb
•GMS973202 CleanWave32Mb

If you look at the ROM that will be used for the Yucatan FX : Re: New Waveblaster card : Reptile Paradise - Yucatan FX and Re: New Waveblaster card : Reptile Paradise - Yucatan FX you can see it is a GMS963200

If you look at the ROM from the Monster MIDI that I have : Re: Roland SCB-55 in 2020 - fair price and how to find one? you can see it is a GMS963200

According to http://www.os2museum.com/wp/of-g-men-and-farmers/ , the Roland sampled version continued to be offered as an option along with the CleanWave version, as mentioned before .

The SAM9407 compatible GS soundbank may or may not be the same as the one in the GMS963200 ROM . It would be nice to be sure .

Can you provide some references regarding this ?

You might be right about the ROM versions and i might be wrong about what triggered the lawsuit. Also the cards i was talking about have no ROM's in them but RAM which the software loads the soundbank into when the card is initialized.

As i find the Dream GS ROMs significantly different from a real SC-55 i get puzzles why Roland would get upset about devices on markets they did not care about.

However the Dream SAM 9407 with the particular 4MB soundfont sounds incredibly close to a SC-55, in a blind test i wouldn't be able to distinguish them, as i tested a Roland CM-300 vs a Guillemot Maxi Sound Lite with the 4MB soundbank and personally i was astonished. It seems almost like the SC-55 ROM was reverse engineered and that soundfont was built from that. I don't have proof it happened this way, but i never heard a soundbank that sounded as close to a Sound Canvas like this one does. Far, far superior to a Cancun FX. In fact i also own a Terratec Maestro 32/96 whose ROM combinations sounds pretty much the same as a Cancun.

I don't have an easy way to make recordings but i can identify the soundbank and let everyone know which one it is.

EDIT: OK so i took a look and found that GSSBK320.94B and BANK4M.94B are exactly the same file. Ran a binary comparison on them and they match. The clue was that both had the exact same file header.
Now, this soundbank as played by the SAM 9407 to me sounds different from the Cancun, and for the better. It also sounds pretty much the same as my CM300.

EDIT2: Here's a Youtube recording someone made:

https://www.youtube.com/watch?v=mhOxHBgJyKc

Take this with a grain of salt. No Roland module will play with the amount of effects this user set his EWS64 to. The sound is too wet, loads of reverb and chorus. I prefer when this soundbank is played on a Guilemot card, as by default the effects level is very close to Roland.

Reply 23 of 37, by carlostex

User metadata
Rank l33t
Rank
l33t
darry wrote on 2020-07-11, 15:40:

I absolutely agree with you that the GS ROM is not that similar to a Sound Canvas .

I am starting to be tempted to get a SAM9407 base card with RAM just to test that soundbank you mention . Please let us know, when you get the chance, which 94B soundbank you are referring to . My guess is the GSSBK320.94B one (from reading other posts, not for having heard it) .

It is the GSSBK320.

I have a Guillemot Maxi Sound Lite and a Terratec EWS64 XL. I have my Terratec listed at amibay because i would very much prefer to get the Guillemot Maxi Sound Home Studio 64 Pro, which has the ESS Audiodrive rather than the Crystal 4236. The former has SC8600 printed on the board and the later has SC8500. The former also includes 2MB of RAM i think, and the later includes the GS ROM we talked about before.

Reply 24 of 37, by darry

User metadata
Rank l33t++
Rank
l33t++
carlostex wrote on 2020-07-11, 15:49:
darry wrote on 2020-07-11, 15:40:

I absolutely agree with you that the GS ROM is not that similar to a Sound Canvas .

I am starting to be tempted to get a SAM9407 base card with RAM just to test that soundbank you mention . Please let us know, when you get the chance, which 94B soundbank you are referring to . My guess is the GSSBK320.94B one (from reading other posts, not for having heard it) .

It is the GSSBK320.

I have a Guillemot Maxi Sound Lite and a Terratec EWS64 XL. I have my Terratec listed at amibay because i would very much prefer to get the Guillemot Maxi Sound Home Studio 64 Pro, which has the ESS Audiodrive rather than the Crystal 4236. The former has SC8600 printed on the board and the later has SC8500. The former also includes 2MB of RAM i think, and the later includes the GS ROM we talked about before.

Thank you. I will have a listen with some headphones .

I wonder if it would be possible to convert a 94B format soundbank into a DXB format one and load the resulting file into a DreamBlaster ?

Reply 25 of 37, by keropi

User metadata
Rank l33t++
Rank
l33t++

Both Cancun and Yucatan use the same 4MB "GS" ROM (the "GMS963200 Roland 32Mbit" one as dream calls it in datasheets)
Turtle beach has their own maskrom with a TB part number but that's the same contents.
SAM9733 has the same exact sound engine as the 9407/9503 just lacks features like external ram usage , it is only able to address up to 32mbit of rom data.

B4gR65z.png

Yucatan has the original maskrom from Dream after the deal so that's why it has the GS logo on it.
Check this video, sounds pretty close to the sc-55 to me: https://www.youtube.com/watch?v=NGog_cqCaus

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 26 of 37, by carlostex

User metadata
Rank l33t
Rank
l33t
darry wrote on 2020-07-11, 16:05:

I wonder if it would be possible to convert a 94B format soundbank into a DXB format one and load the resulting file into a DreamBlaster ?

Well i thought about the same before and asked member dreamblaster about it. It is not that straight forward.

@keropi

Yeah i'm aware about SAM9733 and it being a low cost version of the 9407. But in any case i find the GSSBK320 soundbank superior to the content on the GMS963200 Roland 32Mbit ROM. Leo if you think the ROM sounds close to Sound Canvas you gotta listen to this GSSBK320 soundfont for the 9407. The soundfont will also work on a 9733 of course. The recordings on Youtube don't do the GSSBK320/9407 justice because people put way too much effects by default, i think it might be the default settings from Terratec EWS 64XL. On my Guillemot Maxi Sound Lite 64 the effects level are perfect. I was astonished when i tried the Roland CM300 and Guillemot side by side with my mixer in real time, because they sounded virtually the same. Sometimes speculation is not good because it might create myths, but i don't think speculating that "Roland SC ROM contents were dumped, reverse engineering and built into this sound bank" is far from being a possibility.

They sound so close that i have to assume this was the catalyst for the lawsuit: the combination of the powerful SAM 9407 + GSSBK320 soundbank on a musical instrument could eventually upset Roland's target market. While the GS ROMS on Cancun FX and MAestro 32/96 and others sound different enough and were targeted to markets that Roland wasn't interested in. Sound cards and Karaoke machines. Rich Heimlich mentioned it to me personally and i think he even posted it somewhere here on the forum that they weren't interested in the PC soundcard market at all.

But i might be wrong about all this. But again and personal opinion GSSBK320 + 9407 is superior to any DREAM ROM offerings, including the Cancun FX. I don't mean the latter is bad, i just feel the former is much better.

Reply 27 of 37, by keropi

User metadata
Rank l33t++
Rank
l33t++

Is there a place with a good 9407 + GSSBK320 recording?

If we are to believe the datasheets then 9407-9733 should be the same engine-wise. Ofcourse this does not mean that GSSBK320 and GMS963200 are the same to begin with.
It is my understanding that the lawsuit started because of the GMS931601+GMS931600 roms that later got merged to a single maskrom and renamed to GMS963200 bearing the GS marking.

I cannot see how we can port GSSBK320 to the 9733 , all those Roland/Cleanwave options Dream offered combined both firmware and soundbank to a single maskrom ... without info/conversion tools this might be impossible to make.

🎵 🎧 PCMIDI MPU , OrpheusII , Action Rewind , Megacard and 🎶GoldLib soundcard website

Reply 28 of 37, by yawetaG

User metadata
Rank Oldbie
Rank
Oldbie
carlostex wrote on 2020-07-11, 15:43:

EDIT2: Here's a Youtube recording someone made:

https://www.youtube.com/watch?v=mhOxHBgJyKc

Take this with a grain of salt. No Roland module will play with the amount of effects this user set his EWS64 to. The sound is too wet, loads of reverb and chorus. I prefer when this soundbank is played on a Guilemot card, as by default the effects level is very close to Roland.

Oh, you can pull that off without problem on a Roland module, if you fiddle with the effect settings. A Roland module can sound a lot wetter than what is heard in that video (which is pretty mild, really).

Reply 29 of 37, by Alex-aut

User metadata
Rank Newbie
Rank
Newbie
Jonas-fr wrote on 2020-07-09, 13:36:

Owner of a SC-55 here, I want to shrink down my setup (486 computer hooked to a SC-55 via MIDI) while retaining the SC-55 abilities. I'm currently focusing on getting a 1:1 copy of the SC-55 soundfont (I won't go in the debate between what's the closest soundfont today ,there's some very good clone/reproduction more or less faithful today but that's not the point of this topic). I'm wondering whether it would be possible to extract the exact sounds contained in one (or two) of the chips of a SC-55 to get a soundfont and then flash it on a Dreamblaster X2.

I am also very interested, to follow jonas-fr idea. I already dumped the three SC-55 ROM chips also the four ROM--Chips from the SC-88. In my opinion, these two synths are only PCM-based rompler without LA-synthesis.

The Problem: I loaded every dump into Audacity and all of the ROMS are scrambled. Either the address lines or the data.lines. Next problem: If you take a look at the schematic of SC55 or SC88 (service notes), the normal CPU does not have any "contact" on address-bus/data-bus to the ROM's. Responsible for ROM-readout is the custom FPGA (in case of SC55, it's the IC24 PCM Custom chip).
So the true problem is to get the right "wire-combination" of address- respective data lines. For address lines it would be the faculty of 20! respective faculty of 16!, because the first 4 address-/data lines are not mixed, because in the beginning of the dump data, "ROLAND GSS INTERNAL VER0.20" is readable.
So the is no way in my opinion, to get the correct line-setup without a logic analyzer, which also control on all the "chip-select" lines .
I am working on this since 2 years and I will not give up to uncover this secret.

Last edited by Stiletto on 2020-08-11, 00:58. Edited 1 time in total.

Reply 30 of 37, by Alex-aut

User metadata
Rank Newbie
Rank
Newbie

@jonas-fr: next problem would be, to get the entry-point addresses as well the looping-start, looping-end point addresses for every sample. these addresses are for my opinion stored in the firmware-ROM. At least only an approximation could by done for the TVF's and TVA's. I guess, that the CPU (Hitachi 532) is doing the MIDI stuff and than sends the corresponding addresses to the custom PCM chip.

Reply 31 of 37, by mattw

User metadata
Rank Oldbie
Rank
Oldbie

this topic seems very closely related to what i've just posted here:

About Roland Virtual Sound Canvas 3

if someone wants to reverse-engineer SC-55/88/Pro, I believe the best bet is to disassembler and study of RVIEg01VST.sys, which is "Roland Virtual Sound Canvas Synthesizer Engine" for PC - Roland released that over 20 years ago (some of the files are with 1999 as date of creation).

I don't know why it's not popular, i.e. that there was software equivalent for PC of SC-55, SC-88 and SC-88Pro basically only few years after SC-88Pro was released.

Reply 33 of 37, by ArielAr

User metadata
Rank Newbie
Rank
Newbie
Alex-aut wrote on 2020-08-03, 20:17:

The Problem: I loaded every dump into Audacity and all of the ROMS are scrambled. Either the address lines or the data.lines. Next problem: If you take a look at the schematic of SC55 or SC88 (service notes), the So the is no way in my opinion, to get the correct line-setup without a logic analyzer, which also control on all the "chip-select" lines .
I am working on this since 2 years and I will not give up to uncover this secret.

Hi Alex ! I am curious if you solved the mytery. I see on ebay some simpler logic analyzers with 16 line of input för under 100 USD...so you could solve the mystery without expensive equipment.

Let me know how it goes for you with this project.

Reply 36 of 37, by AppleSauce

User metadata
Rank Oldbie
Rank
Oldbie
Trelokk wrote on 2022-11-12, 12:24:

Is that EmuSC emulator accurate enough by now? They say on Github still that the sound isn't anywhere near the SC-55 yet. Or is this info outdated?

https://github.com/skjelten/emusc

Here's a recording of doom, it sounds pretty decent , it helps that the emulator uses the original roms , it even does stuff like the sc55 mk2 spinny animation,

Filename
doomemusc2.mp3
File size
775.72 KiB
Downloads
39 downloads
File license
Public domain

There's been some updates and it seems like its picked up steam again:
[Implemented "ALL" button in GUI to set synth (global) settings]
[Added support for key shift from both global and part settings]

Sure its probably going to need some more work , but these things take time.
As it is with the virtual display and the options of different synths its looking pretty good already.

Reply 37 of 37, by Trelokk

User metadata
Rank Member
Rank
Member

Well, it's hard to judge it in its entirety from that limited sample you provided, even though that part sounded quite good. Not worse than the SCVA, for sure. However, I am sure that the emulator is still in such an early development stage that it can hardly be recommended for regular usage already.